What are the two primary techniques used to control the structure of an LLM's output in this exercise?	Message prefilling and stop sequences.
What sequence was initially used as both a pre-filled assistant message and a stop sequence?	Three backticks (```).
Why did the initial use of three backticks (```) cause Claude to add a language identifier (like `bash`)?	Because three backticks indicate the start of a markdown code block, and Claude automatically adds a language identifier for syntax highlighting.
How can you prevent Claude from automatically adding a language identifier when using a markdown code block?	By explicitly including the desired language identifier (e.g., `bash`) within the pre-filled message.
Besides using specific characters (like backticks), what else can message pre-filling be used for?	Dramatically guiding Claude on how to structure or format its response.
What strategy was used to prevent Claude from inserting unwanted explanatory comments within the code blocks?	Using the message pre-fill to provide strong instructions, such as "Here are all three commands in a single block without any comments."
What was the main objective of the exercise regarding the final output format?	To receive all three commands in a single response, side-by-side, without any additional commentary or explanation.